Dr. E. Turner on some Atomic Weights. 1 1 1 



' I may add that the preceding analysis agrees closely with 

 that of Berzelius; but I prefer my own result, because my 

 chloride of lead appears to have been purer than the speci- 

 men employed by him, dissolving in water without the slightest 

 residue. It affords an instructive test of the value of the 

 atomic weights current among us. For, supposing 104, 36, 

 and 110 to be the respective equivalents of lead, chlorine, and 

 silver, it follows that 100 of chloride of lead should yield 

 104- '28 parts of chloride of silver, instead of 103*24 as given 

 by experiment. In fact, as will immediately appear, the equi- 

 valent of silver is still more erroneous than those of lead and 

 chlorine. 



Silver. — My first attempts to determine the equivalent of 

 silver were by means of the oxide; but different analyses dis- 

 agreed so widely, that I was obliged to resort to another me- 

 thod. Knowing very nearly the equivalent of chlorine, that 

 of silver may be inferred from the composition of the chloride. 

 According to Dr. Thomson, 100 parts of silver correspond to 

 132'73, according to Berzelius to between 132-75and 132*79, 

 and by my experiments to 132-8. The coincidence is very 

 close, and therefore the principal difference in the equivalent 

 of silver will depend on that of chlorine. If 36 be assumed 

 as the equivalent of chlorine, that of silver is 110; and it is 

 108"08 if 35"'15 be chosen as the equivalent of chlorine. An 

 extremely slight dilTerence in the number for chlorine, such 

 as lies entirely within the ordinary limits of error, would raise 

 the equivalent of silver to 108*1 or rather higher, or depress 

 it to 108. While the matter is uncertain it will be most con- 

 venient to employ the whole number. 



In order, by an independent analysis, to ascertain which of 

 the numbers above mentioned is the more accurate, I prepared 

 some very pure nitrate of silver, kept it for some time in fusion, 

 and converted it into chloride of silver. After repeated ex- 

 periments, I find that 100 of the chloride of silver corresponds 

 to a quantity of fused nitrate, varying from Il8'*544 to 1 18*50. 

 But the theoretic quantity deduced by adopting 110 and 36 

 to represent silver and chlorine is 117*81, which difTers widely 

 from the result of actual experiment; whereas, supposing 

 the equivalent of silver and chlorine to be represented by 108 

 and 35*45, 100 of chloride of silver should correspond to 

 1'18"51 of the fused nitrate. This, then, is additional evidence 

 in favour of the atomic weight of chlorine as above stated. 



Nitrogen. — I have endeavoured to ascertain the equivalent 

 of nitrogen by the analysis of the nitrates of silver and lead, 



1. From the analysis just stated, Iconsider lOOof thccliloride 

 of silver, containing 76*3012 silver, to be equivalent to 1 18*5 of 



